Dataset Viewer
Auto-converted to Parquet Duplicate
codebase_version
string
trossen_subversion
string
robot_type
string
total_episodes
int64
total_frames
int64
total_tasks
int64
total_videos
int64
total_chunks
int64
chunks_size
int64
fps
int64
splits
dict
data_path
string
features
dict
video_path_original
string
tsfile_conversion
dict
v2.1
v1.0
trossen_ai_stationary
66
127,171
3
132
1
1,000
30
{ "train": "0:66" }
data/coffee_making_demo.tsfile
{ "Time": { "dtype": "int64", "shape": [ 1 ], "tsfile_role": "TIME", "unit": "ms" }, "episode_index": { "dtype": "int64", "shape": [ 1 ], "tsfile_role": "TAG" }, "task_index": { "dtype": "int64", "shape": [ 1 ], "tsfile_role": "TAG" }, ...
videos/chunk-{episode_chunk:03d}/{video_key}/episode_{episode_index:06d}.mp4
{ "source_dataset": "allday-technology/coffeeMakingDemo", "source_data_path": "data/chunk-{episode_chunk:03d}/episode_{episode_index:06d}.parquet", "converted_data_path": "data/coffee_making_demo.tsfile", "table_name": "coffee_making_demo", "granularity": "merged", "time_precision": "ms", "time_mapping": ...

Coffee Making Demo (TsFile)

Apache TsFile version of allday-technology/coffeeMakingDemo.

Overview

A LeRobot (v2.1) teleoperation dataset of a coffee-making demonstration recorded on a Trossen AI stationary robot. Each episode is a single-arm (7-DoF) trajectory with synchronized proprioceptive state and action commands at 30 fps.

The dataset covers the following tasks:

  • pick up a coffee capsule

  • pick up an empty cup by the handle

  • Robot: trossen_ai_stationary, single arm, 7 joints (left_joint_0left_joint_6).

  • Episodes: 66 (episode_index 0–65).

  • Frames: 127,171 time-series rows.

  • Tasks: 3 task ids (task_index 0–2; see meta/tasks.jsonl for the instruction text).

  • Sampling rate: 30 fps (from meta/info.json).

Schema (TsFile structure)

All 66 episodes are stored in a single TsFile table; episode_index and task_index are TAG columns (the TsFile device dimension), so one episode is selected with WHERE episode_index=0.

  • Time (INT64, milliseconds) — round(timestamp * 1000); restarts at 0 for each episode, stepping by ~33 ms (30 fps).
  • episode_index (TAG) — source episode id, 0–65.
  • task_index (TAG) — source task id, 0–2; resolve to the instruction text via meta/tasks.jsonl.
  • frame_index (FIELD, INT64) — frame number within the episode.
  • sample_index (FIELD, INT64) — the source global index column, renamed.
  • observation_state_0 … observation_state_6 (FIELD, FLOAT) — robot proprioceptive joint state, flattened from the 7-element observation.state vector (left_joint_0..6).
  • action_0 … action_6 (FIELD, FLOAT) — joint action command, flattened from the 7-element action vector (left_joint_0..6).

The source timestamp column is dropped because it equals Time / 1000 seconds. No other columns or rows are dropped.

Videos

The two camera video streams of the original dataset (observation.images.cam_high, observation.images.cam_left_wrist) are NOT included in this repository. They are large and non–time-series; obtain them from the original dataset: https://huggingface.co/datasets/allday-technology/coffeeMakingDemo (the videos/ directory).

Usage

Read the .tsfile file with the Apache TsFile Java or Python SDK.

Source & license

Downloads last month
57